Fig 1: Identification of Tissue Cell Types: (A) Initial UMAP separated cell types into 14 clusters that were then merged into 12 clusters of 7 distinct cell classes. (B) Cell types were identified using a manual selection of cell type markers and automatic identification based on scores generated by ScType(). Feature maps (relative expression) and violin plots (RNA counts; y-axis) depict specificity for a select marker for each cell type. Fibroblasts (COL1A1+ ), Myeloid cells (FCGR3A+ ), Neutrophils (TGM3+ ), Schwann cells (S100A1+ ), Endothelial (VWF+ ), Smooth muscle/pericytes (ACTA2+ ), and T cells (CD3E+ ), (C) Three markers were used to verify each cell type manually. The relative expression of each is depicted in the dot plot. Fibroblast clusters share higher relative expression of COL1A1, MMP2, and DCN. Dot size represents the percentage of cells with expression of the marker and color intensity depicts the average expression level. (D)The proportions of most cell types were similar across samples. Schwann cells were mostly derived from tumor samples (98.4%). Myeloid cells were enriched in tumor samples (9.30% vs. 3.84%). Fibroblasts represented over 70% of all samples. For this and all future figures, the naming convention is as follows: Fibroblast 1 (F1), Fibroblast 2 (F2), Fibroblast 3 (F3), Fibroblast 4 (F4), Fibroblast 5 (F5), Fibroblast 6 (F6), Myeloid Cells (My), Neutrophils (Nφ), Schwann cells (SC), Endothelial cells (EC), Smooth Muscle/Pericytes (SM), T Cells (TC).